Estimation of origin-destination matrices by fusing detector data and Floating Car Data

Abstract

Abstract Neither traffic detector counts nor Floating Car Data are able to determine the absolute traffic demand with sufficient accuracy. Therefore, traffic management depends on simple and often inaccurate estimations. In recent years, various research projects have shown that plausible results can be achieved from mere road detector data by applying the Information Minimization model and eliminating redundant information. However, a problem that has not yet been solved in this approach is the lack of information about the structure of traffic demand. This can be extracted from Floating Car Data. To merge both pieces of information into one consistent model the present paper introduces a methodology based on the work of Pohlmann utilizing the Information Minimization Model by Van Zuylen.